So, Prof Davidson, if you don't believe in standard forms of grading anymore, what do you believe in? That's the question I'm asked all the time now. And here are some answers.
I'm supplying these courtesy of questions sent to me by a student from Emory University who is majoring in Educational Studies and Sociology and who is working on a senior capstone project on assessment.
Q: What is your philosophy of grading. How does this play out in your grading and assessment methods?
A: Grading was invented in the late nineteenth century as an expediency. It's an odd concept, when you think about it, reducing all the different ways students master subject matter to one letter or numerical grade. We know that, beyond the classroom, an A student is no more likely to succeed in the world than a B student. Among entrepreneurs, A students are rare. We have spent so much time earning good grades, being tested, working for grades, basing educational success on a grade point average, that we have forgotten that efficiency is not the real motivation for learning. Grading should be a means toward the end of learning. When a grade is an end in itself, we have a debased form of measuring that has little to do with learning.
I now use a simple contract grading system coupled with peer-assessment. Contract grading lays out absolutely all of the requirements of a course and says a student can maturely decide what tasks he or she wishes to take on and then decide on a grade, in advance, depending on those. If there are weekly 500-word blog posts, one multimedia presentation, one public contribution to knowledge, and one-student led unit (building a syllabus, preparing the reference guide, and leading the class on a topic) for an A, then a student signs a contract saying s/he will do all of those things. The contract also says the penalty: such as, if a student misses more than two blogs in the course of the semester, the full course grade automatically drops one half grade point (from 4.0 to 3.5). In other words, penalties for not meeting contractual obligations are severe and automatic.
The second part of this is peer-to-peer learning and peer-assessment. The peer-assessment comes in as a quality control. Each week, two students lead the class discussions and set the readings. They also read all of their classmates blog entries and give feedback on all of them. They determine if the posts are satisfactory, establishing their own criteria for what is or is not satisfactory and explaining those. Their goal is to make sure each blog post is satisfactory so, if one is not, they work with their classmate to improve it. They then register on the class contract whether or not the student has succeeded. The next week, those two student leaders are back in the class and two other students lead the class and comment on the work of other students. It is a fantastic learn-by-doing method that turns assessment into a process of learning--for everyone.
Q: What made you try self-assessment? Were there any specific articles or influences?
I have spent the last few years writing a book, now in galleys and in the bookstores next June, called Now You See It: How the Brain Science of Attention Will Transform the Way We Live, Work, and Learn (Viking Press). One chapter of the book is called "How We Measure" and it capsulizes about a decades' worth of research on how we came to grade, measure, and test the way we do now. There is nothing "natural" about grading. It is an invented system for a specific historical moment. That historical moment has changed but we have entrenched an antiquated system throughout our educational system. Our national educational policy, No Child Left Behind, is based on item-response end-of-grade testing and, for my money, it's a nightmare. (Here's the book url for the preorder for anyone who is interested: http://www.amazon.com/Now-You-See-Attention-Transform/dp/0670022829/ref=...)
Q:What are the upsides and downsides of grading and assessing?
The upside is we need some way of telling if we are doing well or badly and, as humans, we are not very good at seeing our own qualities very well. External measures help us to evaluate what we cannot see about ourselves. It is also useful to know how one community (a class, a school, a neighborhood, a city, a region, a country) is doing relative to others--if that comparative research then leads to the social and educational changes that will improve everyone's abilities. The downside is that present methods do not do that. They give us not so much a picture of ourselves but a blurry snapshot that puts the wrong object in the center. I am encouraged by new computational methods that are based on game mechanics and that allow students to take not one end of grade test but tests for everything all the time, organically and with no consequences, and then immediately offer not only feedback but suggestions and, in some cases, automatically offer new challenges based on how well one did or did not do in the past one. Calibrated, just-in-time assessment is about learning. And it can be fun, not demoralizing or frightening.
I predict that in a decade this will be the dominate mode of testing. Even the SAT now will give students feedback based on their scores about the kinds of questions they do well on and so forth. Already that is an improvement over the raw number approach. It is a small step towards geniunely productive assessment that assumes that the grade is not the end of the learning; a grade is a path on the way to learning that never ends.
Bonus Question (I supplied this one): What is a better way of assessing real learning than the current system of grading?
I very much like the kinds of guidelines for learning and working together that the Mozilla Foundation is pioneering for its open source developer community. You don't get a good grade when you are contributing to web development projects. If you do a good job, you are welcomed and supported by the community for your contribution. It's not all touchy-feeling. Quite the opposite. It is about producing, about making your ideas work through clarity and communication, and in constant feedback from others. Matt Thompson of the Mozilla Foundation has recently put together a marvelous "5 Steps for Drumbeat" as part of the Mozilla Drumbeat developer strategy. I highly recommend it as a paradigm shift from hierarchical learning where an expert--a teacher--issues an assignment, a student does the assignment, and then the teacher grades the assignment. Is that learning? Or is that learning how to do well on a topic assigned and graded by a teacher?
Next semester, in both of my classes--"This Is Your Brain on the Internet" and "Twenty-First Century Literacies"--I will be using Matt's method, testing it and using my class to give him feedback. Here's a summary and the link. As is the open source way, he is accepting feedback from anyone who has feedback to offer because the objective in peer-learning isn't the grade you get. The objective is how you can constantly improve and grow. To me, that is great teaching because that is great learning.
* * *
5 Steps for Drumbeat by Matt Thompson, Mozilla Foundation http://openmatt.wordpress.com/2010/11/30/5-dance-steps-for-drumbeat/#com...
- Strategy whats your big idea?
- Story help people understand it.
- Tools set up simple tools that make it easy for contributors to see whats happening and get involved.
- People who are you trying to reach? how can they help right now?
- Prototype build fast. test and improve it together.
(Shake and repeat.)